Proposing a regional gender equality index (R-GEI) with an application to Italy

ABSTRACT

Gender equality represents a central topic of our society, and its study is gaining increasing attention in the international panorama. During the last 20 years, various indicators aiming at measuring gender equality have been proposed, but there are no systematic experiences of indicators tailored for a subnational analysis. We propose a regionalization of the most complete and detailed gender equality indicator, the gender equality index (GEI) of the European Institute on Gender Equality, choosing Italy as a case study. The results show how a regionalized approach to gender equality is necessary to set priorities and target regional policy actions.

KEYWORDS:

• equality
• gender
• gender equality index
• Italy
• regional studies

JEL:

• C4
• D63
• J16
• P48

DISCLOSURE STATEMENT

No potential conflict of interest was reported by the authors.

Notes

1. We prefer to discuss gender equality rather than gender inequality consistently with what is done in primary social contexts (e.g., growth, social trust, sustainable development), with UN Sustainable Development Goal 5 ‘Gender Equality’ and the EIGE’s gender equality index that is the primary motivating example of this work.

2. The first EIGE report (EIGE, 2017a) considered gender equality in the EU-28 countries for 2005, 2010, 2012 and 2015. The 2019 set of national and thematic reports, and the online dashboard, consider GEI scores for 2005, 2013, 2015, 2017 and 2019.

3. The indicators derivable from the gender statistics databases were obtained from the database Women and Men in Decision Making, maintained by the EIGE in its Gender Statistics Database.

4. The geometric mean is more appropriate than the arithmetic mean to provide syntheses of indicators being more robust against outliers and to the compensation effect (Maggino, 2015).

5. The three single years results are not provided in the main text of this paper, but see Appendix A in the supplemental data online.

6. Variables correlated with PC1 and PC2 are the most important in explaining the variability in the data set. The contributions of a PC are computed as the squared loadings for each variable divided by the overall sum of the squared loadings for that component. If the contribution of the variables were uniform, the expected value would be 1/6 = 0.167 (i.e., 1/number of indicators). The dashed line in Figure 3(b, c) indicates the expected average contribution. For a given component, a variable with a contribution larger than this cut-off could be considered as important when contributing to the component.